August 29, 2004

BS conditional semantics and the Pinocchio effect


Wow! According to this World tribune news bulletin, which cites a Middle East Newsline report, a new Israeli product can "determine with 98 percent accuracy whether a suspicious traveler has intentions to launch an attack during flight." And it's based on a voice analyzer in a chip small enough to fit in the frame of a pair of glasses, a chip which switches a light on for the user, in realtime of course, to signal (dis-)honesty. The technology was developed by a mathematician by the name of Liberman (hmm, that name sounds familiar, was he the one who co-discovered contradiction contours?), and is manufactured by a company he founded called Nemesysco.

Where to start with a story like this? How about the 98% claim? I will now demonstrate for you a voice analysis system I designed earlier today which, while still at the beta test stage, appears to surpass the accuracy of the Israeli model by close to two percentage points.

I won't discuss the technology involved in your correspondent's system, save to reveal that the core technology is a famously impenetrable dictum of the logician Iksrat der Flard, who pioneered BS conditional semantics:

"Snow is white" is a lie iff the speaker knows damn well it ain't so.

Having thus established the base case, the remainder of the recursion falls out naturally, and we merely need to add the basic intonation recognition capability for detecting BS at the level of individual propositions. Fortunately, your browser has these capabilities built in, and it was a simple matter to engineer the system, which I now present.

For each of the following three sentences, read the sentence aloud in a normal speaking voice, and then click on the sentence to see the system's honesty analysis:





Isn't technology amazing? Just to check that the system really performs at close to 100%, I'm keeping a running tally of system successes and failures. Hit one of the following two buttons (preferably while reading aloud what it says) to indicate the accuracy of the system's analysis. Needless to say, I don't in any way depend on your honesty in providing this feedback...



So much for 98%. Let's move on to the technology.

For all I know, conventional polygraph technology probably indicates little more than that people who are tense often sweat. So the bar may be low in the lie detection industry. Nemesysco's systems may or may not beat standard polygraphs (although they are certainly less intrusive), and may or may not be useful. Independently of what the technology actually is, the company's approach to presenting its technology suggests their main goal is to pull the wool over the eyes of the public, terrorists and governmental or other customers alike. In the US, Nemesysco systems have previously been marketed with the slogan “The DNA of thought.” Ehh, pardon? Many recent media references to the company's products mention that the analysis performed by the systems requires "8000 algorithms." Gosh, what a lot of algorithms... but heck, from a mathematical perspective I use at least 8000 algorithms every time I flush the toilet. The company's website leaves us none the wiser as to what the technology is really about. It spouts all sorts of nonsense like the following:

The LVA uses a patented and unique technology to detect "Brain activity finger prints" using the voice as a "medium" to the brain and analyzes the complete emotional structure of your subject. Using wide range spectrum analysis and micro-changes in the speech waveform itself (not micro tremors!) we can learn about any anomaly in the brain activity, and furthermore, classify it accordingly. Stress ("fight or flight" paradigm) is only a small part of this emotional structure… 

So what's the big deal here? What for that matter, is the business model? Well, it may be simple. Nemesysco's biggest customers appear to be insurance companies. They buy Nemesysco's business solutions for assessing insurance claims. The insurance companies have nothing to lose, in the sense that they start off with no sensible way of telling whether claims are fraudulent, or which parts of claims are fraudulent. But they do know that if the claimants think that the insurance companies have a way of telling truth from fiction, then fewer fraudulent claims will made or sustained. Indeed, Nemesysco's website says that they offer training courses for insurance claim processors in how to politely indicate to a claimant that the falsity of a claim has been discovered, in the hopes that the claim will then be dropped. The insurance company has little interest, then, in whether Nemesysco's software really works. For what Nemesysco is really selling is is a great patter.

What we have here, I believe, is a technologically updated new release of that psychologically sophisticated version of the Pinocchio effect that evil parents have been using on their innocent children for generations, an effect whereby one lie spawns another, and all in the cause of establishing a norm of honesty:

Mom: Have you washed your hands.
Kid: Yes, Mom.
Mom: I can see your nose growing...

Posted by David Beaver at August 29, 2004 04:52 AM